Jon will be presenting some of the latest and greatest on using linguistically reasonable models of incremental parsing to probe EEG signals. This project , a collaboration with John Hale, Chris Dyer, and Adhi Kuncoro, uses Chris’ RNNG grammar to derive predictions about processing complexity and uncertainty. One thing that is really exciting about this project is John’s implementation of beam search with the RNNG; unlike our earlier incremental-parsing-meets-brains papers, this one models uncertainty and parallel search!
The talk will be at the CRCNS meeting in Berkeley June 13-14 (time TBD).